Since medical procedures and chemotherapy cause major problems for a majority of children with cancer, even producing noncompliance with treatment for some, the objectives of this study are: (1) to determine if hypnosis is more effective than nonhypnotic intervention in reducing acute pain and anxiety related to medical procedures in children; (2) to compare hypnosis, supportive counseling, and assessment alone (control group) in their effectiveness in reducing nausea and vomiting in children and adolescents receiving cancer chemotherapy; (3) to determine if symptom reductions achieved with intervention are maintained after termination of intervention; and (4) to determine if hypnotic susceptibility is related to the degree of symptom reduction for patients receiving hypnosis. Pain and anxiety will be assessed in children (5-10 years) using patient (pt.) and observer (obs.) ratings and a behavioral checklist during 2 bone marrow aspirations and/or 2 lumbar punctures prior to intervention. Pts. (45-60) will be randomized to a hypnosis, supportive counseling, or control group and 2 more procedures will be assessed. For the second study, the duration (hours) and severity (pt. and parent ratings) of nausea (N), vomiting (V), and the extent to which these symptoms bother children (ages 5-21 years) will be assessed within 3-5 days after each of 2 courses of chemotherapy. Forty-five to 60 patients with N and V will be randomized to a hypnosis, supportive counseling, or control (assessment alone) group and the above data will be obtained during 4 more courses (matched with baseline courses for each pt.); pts. in the first 2 groups will receive intervention for the 2 courses after baseline. Mean pt. ratings of N, V, and bother for each time period for the 3 comparison groups will be analyzed by 3 x 3 ANOVA with repeated measures for time. A Friedman ANOVA will test the null hypothesis that no changes in duration of symptoms occurred over time. Comparisons across groups for each of the 3 time periods will be made using Kruskal-Wallis ANOVA's. Reliability checks on pt. ratings will be made by comparisons (Pearson's r) with parent ratings. Hypnotic susceptibility scores will be divided at the median and compared (2 x 2 chi-square) to change scores for pain and anxiety and for N and V. The overall significance of this study is to decrease the suffering of children with cancer and perhaps to improve survival by decreasing noncompliance with treatment.